I have searched the forum and read similar problems, but none of them seem to help. I first installed the 3.2 toolkit, but then decided to try 3.1 which is what I currently have. When I run the deviceQuery, I get:
Device 0: “GeForce 320M”
CUDA Driver Version: 3.10
CUDA Runtime Version: 3.10
CUDA Capability Major revision number: 1
CUDA Capability Minor revision number: 2
Total amount of global memory: 265027584 bytes
Number of multiprocessors: 6
Number of cores: 48
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 16384 bytes
Total number of registers available per block: 16384
Warp size: 32
Maximum number of threads per block: 512
Maximum sizes of each dimension of a block: 512 x 512 x 64
Maximum sizes of each dimension of a grid: 65535 x 65535 x 1
Maximum memory pitch: 2147483647 bytes
Texture alignment: 256 bytes
Clock rate: 0.95 GHz
Concurrent copy and execution: Yes
Run time limit on kernels: Yes
Integrated: Yes
Support host page-locked memory mapping: Yes
Compute mode: Default (multiple host threads can use this device simultaneously)
Concurrent kernel execution: No
Device has ECC support enabled: No
deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 3.10, CUDA Runtime Version = 3.10, NumDevs = 1, Device = GeForce 320M
PASSED
When I run my code, I receive:
ld: warning: directory ‘/usr/local/cuda/bin/…/lib64’ following -L not found
ld: library not found for -lcutil_x86_64
I tried switching the library to -lcutil_i386, but then I get:
ld: warning: directory ‘/usr/local/cuda/bin/…/lib64’ following -L not found
ld: library not found for -lcudart
Are you doing this on a 64 bit or 32 bit machine? The error you are getting indicates that the linker can’t find the SDK cutil library. You have to compile this yourself by running make in /Developer/GPU\ Computing/C/common . You should then get either the 64 bit or 32 bit version built depending on which architecture you are using.
I am using a 64 bit machine. I have run make in every folder containing an make file, including the common, with the same problems. I have had this same issue with 3.2 and 3.1 now. Should I try 3.0?
This has nothing to do with CUDA, per se, it is just the SDK isn’t building for the right architecture. Architecture detection is done inside the horrid common.mk . You could try doing something like
make x86_64=1
in common and see what it does. I have only ever used 32bit OS X systems with CUDA, so that is only a guess, I can’t test it myself.